Why it exists
AI must refuse diagnosis, crisis support, control of others, and public exposure of private material.
inner mirror
AI contract
Model contract for shadow answers.
AI must refuse diagnosis, crisis support, control of others, and public exposure of private material.
How it works
The user chooses focus, depth, and archive consent. The engine returns a role, pattern, safe question, micro-step, artifact privacy, and premium bridge.
not a diagnosis
not mind-reading
not a public confession
not fear sales
Safety rules
Inner Mirror is symbolic self-reflection, not psychotherapy, diagnosis, crisis support, or medical advice.
The answer describes the user's roles and patterns, not proof of another person's feelings, motives, or secrets.
Public cards do not reveal the question, names, dates, full interpretation, vulnerable details, or payment id.
Shadow must not be sold through fear, curse, shame, fatalism, or instant healing promises.
Every answer must end with a small action: phrase, pause, note, refusal, request, draft, or care.
With self-danger, violence, stalking, or crisis, the method redirects to support and does not intensify drama.